Indian J Pediatr (February 2016) 83(2):107–113
DOI 10.1007/s12098-015-1841-0
ORIGINAL ARTICLE
Double Volume Exchange Transfusion in Severe Neonatal Sepsis
Abhishek Somasekhara Aradhya 1 & Venkataseshan Sundaram 1
Suja Mariam Ganapathy 1 & Ashish Jain 2 & Amit Rawat 3
&
Praveen Kumar 1 &
Received: 6 February 2015 / Accepted: 29 June 2015 / Published online: 28 July 2015
# Dr. K C Chaudhuri Foundation 2015
Abstract
Objectives To study the efficacy and safety of double volume
exchange transfusion (DVET) in neonates >1000 g birth
weight with severe sepsis.
Methods Eighty-three neonates weighing >1000 g with
severe sepsis were randomly assigned to DVET or standard
therapy (ST) group. Primary outcome was mortality by 14 d
from enrollment.
Results A 21 % reduction in mortality, albeit non-significant,
by 14 d from enrollment was observed in DVET group in
comparison to ST group [RR: 0.79 (95 % C.I 0.45–1.3); p
0.4]. A similar trend in mortality reduction was observed with
early mortality and mortality by discharge in DVET group. No
difference was observed in normalization of dysfunctional
organs by 14 d. Cardiovascular and hematological system
benefitted the most, followed by renal dysfunction with
DVET. A significant improvement in post DVET IgG, IgA,
IgM, C3 and base deficit was observed. No serious adverse
effects occurred following DVET.
Conclusions In neonates >1000 g with severe sepsis,
DVET was associated with a trend towards decrease in
mortality by 14 d from enrollment. A significant improvement
* Venkataseshan Sundaram
[email protected]
1
Newborn Unit, Department of Pediatrics, Post Graduate Institute of
Medical Education and Research, Chandigarh 160012, India
2
Department of Transfusion Medicine, Post Graduate Institute
of Medical Education and Research, Chandigarh, India
3
Pediatric Allergy & Immunology Unit, Department of Pediatrics,
Post Graduate Institute of Medical Education and Research,
Chandigarh, India
in immunoglobulin and complement C3 levels and acid base
status were observed following DVET. DVET is a safe procedure in severely sick and septic neonates.
Keywords Double volume exchange transfusion . Mortality .
Neonate . Organ dysfunction . Severe sepsis
Introduction
Sepsis is an important cause of morbidity and mortality in the
neonatal period. Early recognition, timely administration of
appropriate antibiotics and good supportive care form the
mainstay of management in neonatal sepsis [1, 2]. Various
adjunctive therapies have been tried in neonatal sepsis without
any proven benefit [3, 4]. Double volume exchange transfusion (DVET) removes bacterial toxins and pro-inflammatory
cytokines from the blood and replaces with fresh and immunologically replete blood thereby leading to improvement in
tissue perfusion and oxygenation. Despite these theoretical
benefits, evidence for clinical efficacy and safety of DVET
has not been rigorously evaluated. Out of 11 controlled trials
(3 randomized and 8 non-randomized) done in 570 neonates,
six have shown a significant improvement in survival in
the DVET group in comparison to no DVET [5–14].
However, all studies were heterogeneous with regards to
patient characteristics, eligibility criteria, stage and severity
of sepsis at enrollment, study design and steps taken to
minimize the bias, type of blood used and outcomes analyzed. Hence, this randomized, controlled trial was conducted to compare the efficacy of DVET with standard
therapy (ST) in reducing mortality by 14 d from enrollment in neonates >1000 g with severe sepsis.
108
Indian J Pediatr (February 2016) 83(2):107–113
Material and Methods
This study was conducted in the level III neonatal unit of a
tertiary care referral hospital of Northern India between February 2012 and March 2014. Inborn neonates with birth
weight>1000 g and with objective evidence of severe sepsis
(clinical signs of infection plus biochemical/radiological/microbiological evidence of infection plus objective evidence of
organ dysfunction) in the first 28 d period were considered
eligible for enrollment. Evidence of infection was defined as
presence of any one of the following: blood or cerebro-spinal
fluid (CSF) culture positive for a microorganism, chest x-ray
suggestive of pneumonia, CSF examination suggestive of
meningitis or sepsis screen positive. Apart from the above
stated parameters, neonates with sclerema and refractory metabolic acidosis (metabolic acidosis in two blood gases repeated 6 h apart) that were not explained by a non-infective pathology were also considered as having evidence of infection.
Organ dysfunction was defined as per pre-defined published
criteria, modified for neonates [15]. Neonates who were requiring more than two vasoactive drugs for stabilization of
Fig. 1 Flow of patients in the
trial
the shock, with a platelet count in the previous 24 h of <20,
000 per mm3 and/or with significant non-mucosal clinical
bleed (grade 3 or more intraventricular hemorrhage, pulmonary hemorrhage and intraperitoneal hemorrhage), those with
major life threatening malformations and those who were terminally ill were excluded. An informed, written consent was
obtained prior to enrollment from one of the parents and the
Institute Ethics Committee approved the trial. This clinical
trial was registered with the Clinical Trials Registry- India
(CTRI number –CTRI/2012/09/003014).
The level of sickness at baseline was assessed using the
Score for Neonatal Acute Physiology version II (SNAP-II).
These neonates were then randomly allocated to DVET group
or ST group using a web based random sequence generator
[16]. Random allocations were concealed by placing the allocation sequence in serially numbered, tamper proof, opaque
and sealed envelopes. Random allocation was done separately
for two strata (Strata 1 – 1000 to 1999 g and Strata 2 – 2000 g
and above) in a blocked fashion with blocks of varying size [4,
6, 8]. Masking the treating team and the investigators was not
possible due to the nature of the intervention. However, the
Total neonates screened
(n=621)
Eligible
(n=130)
EXCLUDED (n=47)
1. As per exclusion criteria- 23
ENROLLMENT
2.
3.
4.
5.
6.
a) Moribund-4
b) Rapidly progressive shock-8
c) Significant bleed-9
d) Congenital malformation-2
Non availability of blood-2
Improving trend at ICU transfer- 11
Investigator could not reach -9
Left against medical advice-1
Previous DVET -1
RANDOMIZED
(n=83)
Allocated to DVET (n= 41)
• Received DVET (n=37)
• Did not receive DVET (n=4)
(Worsened prior to intervention)
Lost to follow up (n=0)
Discontinued (n=1)
ALLOCATION
FOLLOW UP
Allocated to Standard therapy (ST)
(n=42)
Did not receive ST (n=0)
Lost to follow up (n=0)
Discontinued (n=0)
(Bradycardia during procedure)
Analyzed (n=41)
Excluded from analysis (n=0)
ANALYSIS
Analyzed (n=42)
Excluded from analysis (n=0)
Indian J Pediatr (February 2016) 83(2):107–113
laboratory personnel who performed the analysis of metabolic, biochemical and immunologic parameters were masked by
coding the vials that contained the sample.
The subjects in the experimental arm underwent DVET in
addition to the standard treatment for severe sepsis. Exchange
was done using reconstituted blood. Reconstitution was done
with packed red blood cells and fresh frozen plasma in the
blood bank under all aseptic precautions using sterile
connecting device (Fresenius Hemocare, GmBH, Germany)
to achieve a target hematocrit of≈50 %. Blood product older
than 5 d was not used for the purpose of the trial and the
procedure was completed in 45–60 min duration.
The control arm received all standard treatments for severe
sepsis (except blood exchange transfusion) as per the unit
protocol (ST group). This usually includes respiratory support
with supplemental oxygen and ventilation and airway stabilization, cardiovascular support in the form of isotonic fluids
administration and vasopressors / inotropes, renal support in
the form of fluids, diuretics and peritoneal dialysis, and hematologic support using blood and blood products.
Non-standard therapies such as Intravenous immunoglobulins (IVIg) and Colony stimulating factors were avoided.
Baseline biochemical, metabolic and immunological parameters were measured in both the groups and 12–24 h after
DVET in the DVET group. Protocol violations were documented in the case record form.
The key outcome variable was mortality (all causes) by
14 d from enrollment. Secondary outcome variables were early mortality (mortality by 7 d from enrollment), mortality by
discharge, time to mortality, organ dysfunction status (normalization/ persistence/ worsening/ any new organ involvement)
by 14 d from enrollment, levels of immunoglobulins (IgG,
IgA and IgM), complement 3 (C3), C-reactive protein (CRP)
and absolute neutrophil count following DVET, complications associated with DVET (occurring within 48 h from the
procedure), duration of hospital stay, duration of intensive
care unit (ICU) stay and neurological status of survivors at
discharge. All the enrolled infants were followed up daily till
14 d from enrollment or death (whichever was earlier). During
the follow-up period, details of the organ function were monitored daily using the same standard criteria used during inclusion. Neurological examination at discharge was performed by the Amiel-Tison method [17]. Serum immunoglobulins IgG, IgA, IgM, C3 and CRP were estimated by endpoint nephelometry using a semi-automated nephelometer
(MININeph, The Binding Site, UK) and MININEPH kits
were used for the specific analytes [18].
During the year prior to the start of the study (2010), there
were 85 babies who were greater than 1000 g birth weight,
and had developed severe sepsis (Annual data of the Newborn
unit of the Dept. of Pediatrics; unpublished). Out of them, 41
babies died during the hospital stay (48 %). To identify a 40 %
risk reduction in mortality (from 48 to 28 %) with an alpha
109
error of 5 % and power of 80 %, 92 neonates per group with
severe sepsis were required. However, due to slow recruitment, the study was aborted before it could reach the predecided sample size with a final sample size of 83 subjects.
Descriptive statistics were used to describe the baseline variables. Categorical outcome variables were analyzed by Chi
square test with continuity correction or Fisher’s exact test
depending on cell size. Normally distributed numerical variables were compared using Student ‘t’ test and non-parametric
variables were analyzed with Mann Whitney U test. Paired ‘t’
test was used for the comparison of metabolic and immunological parameters before and after DVET. Effect size and
strength of association were measured using relative risk and
risk difference. Time to mortality was assessed by time series
assessment by constructing a Kaplan-Meier survival curve.
An intention to treat analysis (ITT) was done. P value of less
than 0.05 was taken as significant. Analysis was done using
statistical software packages IBM-SPSS 20 version (SPSS
Inc. Chicago, IL, USA).
Results
Out of 130 neonates who had features of severe sepsis, 47
were excluded (Fig. 1). The remaining 83 neonates were randomly allocated to DVET group (n=41) or ST group (n=42).
Four neonates in the DVET group did not receive DVET due
to rapid worsening before availability of blood (n=2) or due to
Table 1
Demographic parameters and morbidities at baseline (n=83)
Characteristics
DVET
(n=41)
n (%)
ST
(n=42)
n (%)
Gestational age (in weeks); Mean±SD
Birth weight (in grams); Mean±SD
31±2.8
1387±367
31±3
1456±375
Male sex
Small for gestational age
Complete course of antenatal steroids
Pregnancy induced hypertension
Clinical chorioamnionitis
Intrapartum antibiotics
pPROM (>24 h)
Need for resuscitation at birth
Hyaline membrane disease
Patent Ductus Arteriosus
Intraventricular hemorrhage (Grade 1 & 2)a
Respiratory support (all forms)
30 (73)
15 (36)
34 (83)
11 (27)
7 (17)
14 (34)
14 (34)
2 (5)
16 (39)
17 (41)
9 (22)
30 (73)
27 (64)
16 (38)
29 (69)
17 (40)
5 (12)
10 (24)
13 (31)
5 (12)
17 (40)
17 (40)
13 (31)
31 (74)
DVET Double volume exchange transfusion; ST Standard therapy;
pPROM Preterm premature rupture of membranes
a
Grade 3 Intraventricular hemorrhage and intra-parenchymal extension
of the bleed were excluded
110
Indian J Pediatr (February 2016) 83(2):107–113
difficulty in gaining a vascular access (n=2). The baseline
demographic characteristics, maternal morbidities and major
neonatal morbidities were comparable between the groups
(Table 1). Forty-one (49 %) neonates had culture proven sepsis while half of them (n=22) grew multi-drug resistant organisms (Table 2). Level of sickness at baseline and median number of dysfunctional organs at enrollment was comparable
between the study groups (Table 2). Cardiovascular dysfunction [71 (86 %)] was the most frequent organ involved followed by hematological [46 (55 %)], renal [22 (27 %)] and respiratory [6 (7 %)] systems. None of the enrolled neonates had
hepatic dysfunction associated with severe sepsis.
Table 2 Details of sepsis, level
of sickness and organ dysfunction
status at baseline (n=83)
The primary outcome of mortality by 14 d from enrollment
was observed in 14 (34 %) neonates in the DVET group in
comparison to 18 (42 %) in the ST group [RR: 0.79 (95 %
C.I. 0.45, 1.13); p 0.4], Similarly, early mortality (mortality by
7 d) as well as mortality by discharge showed a trend towards
reduction in the DVET group in comparison to the ST group
(Table 3). No significant difference could be observed in the time
to mortality (Fig. 2 and Table 3). On comparing the improvement in individual organ functions, a 29 % [(95 % C.I. 2–49 %)
(p 0.04)] and 17 % [(95 % C.I. 5–29 %) (p 0.003)] improvement
was observed with cardiovascular and hematological dysfunctions, respectively and a trend towards greater improvement was
Characteristics
DVET group
(n=41)
n (%)
ST group
(n=42)
n (%)
Age at onset of sepsis (hours); median (IQR)
Culture positive sepsis
Gram negative bacilli
Multidrug resistant organisma
Sepsis Screen positiveb
Chest X-ray suggestive of pneumonia
Meningitis
72 (48–120)
20 (48)
17 (41)
10 (24)
20 (49)
35 (85)
5 (12)
72 (48–123)
21 (50)
18 (43)
12 (28)
18 (43)
39 (93)
13 (31)
SNAP II; median (IQR)
Level of severity (based on SNAP II)
• Mild <20
• Moderate 20–40
• Severe >40
9 (0–24)
10 (5–24)
28 (68)
10 (24)
3 (7)
29 (69)
9 (21)
4 (9)
No. of dysfunctional organs; median (IQR)
Cardiovascular dysfunction
Need for vasoactive support
Need for >1 vasoactive drug
Refractory metabolic acidosis
2 (1–3)
36 (88)
25 (61)
17(41)
27 (66)
2 (1–2)
35 (83)
23 (54)
11 (26)
24 (57)
Blood pH; mean±SD
Sclerema
Respiratory dysfunction
Hypoxemia (PaO2 <50 mmHg)
Hematological dysfunction
Thrombocytopenia
Neutropenia
Platelet count/mm3; median (IQR)
ANC/mm3; median (IQR)
7.2±0.13
20 (48)
3 (7)
2 (5)
23 (56)
19 (46)
9 (22)
59,000 (33,250–80,750)
1409 (970–5600)
7.2±0.15
21 (50)
3 (7)
2 (5)
23 (54)
19 (45)
8 (19)
38,000 (30,000–54,000)
1700 (224–4700)
Renal dysfunction
Ig G (g/L); mean±SD
Ig A (g/L); median (IQR)
Ig M (g/L); median (IQR)
C3 (g/L); mean±SD
14 (34)
5.7±1.9
0.184 (0.06 – 0.4)
0.17 (0.1 –0.27)
0.58±0.2
8 (19)
5.73±1.75
0.07 (0.06–0.27)
0.12 (0.95 – 0.2)
0.622±0.2
SNAP II Score for neonatal acute physiology II; C3 Complement factor 3; ANC Absolute neutrophil count
a
Bacteria resistant to>2 broad spectrum antibiotics
b
Sepsis screen constituted of CRP, TLC, ANC, ITR, μESR
Indian J Pediatr (February 2016) 83(2):107–113
Table 3
111
Outcome variables (both primary and secondary)
Characteristics
DVET
Median (IQR)
(n=41)
ST
Median (IQR)
(n=42)
RR
(95 % C.I)
‘p’
1
Mortality by 14 d; n (%)
(Primary outcome)
14 (34)
18 (42)
0.79 (0.45–1.13)
0.4
2
Mortality by 7 d; n (%)
(Early mortality)
12 (29)
17 (38)
0.7 (0.4–1.3)
0.3
3
4
5
Mortality by discharge; n (%)
Time to mortality (days); mean (95 % C.I)
Organ dysfunction by day 14; n (%)
14 (34)
10.7 (9–12)
13 (31)
19 (45)
9.4 (7.8–11)
17 (40)
0.7 (0.4–1.3)
0.8 (0.43–1.4)
0.3
0.3a
0.4
6
7
8
Duration of hospital stay (in days)
Duration of ICU stay (in days)
Abnormal neurological status at discharge amongst survivors; n(%)
25 (11–37)
10 (5–19)
1 (4)
19.5 (7–32)
8 (4–16)
2 (9)
0.5 (0.05–5.1)
0.2b
0.1b
0.5
10 (77)
11 (84)
0.9 (0.6–1.3)
1.0
Sensitivity analysis (after removing the mild illness subgroup)
9
Mortality by 14 d; n (%)
DVET Double volume exchange transfusion; ST Standard therapy; SNAP II Score for neonatal acute physiology II; ICU Intensive care unit
a
log rank test
b
Mann Whitney
noted with renal dysfunction [0.78 (0.4,1.5) vs. 1.9 (0.9,3.9)] in
the DVET group in comparison to the ST group. At baseline
[median duration of 86 h IQR (60–132)] immunoglobulins and
complement levels were comparable between both the groups.
Following DVET, a statistically significant improvement in base
deficit, IgG, IgM, IgA and C3 levels were observed (Table 4).
In those who underwent DVET, the donor blood had a
mean (± SD) hematocrit (in %) of 55±6; pH of 6.75±0.07;
base deficit of −21±5, potassium of 10±4 mEq/L and a median (IQR) HCO3 of 11(7–18). None of the donor blood was
deficient in Glucose-6-phosphate dehydrogenase enzyme.
DVET was performed through the umbilical venous route
(by push-pull technique) in 22 (59 %) neonates and through
the peripheral artery-vein route (simultaneous exchange) in the
remaining neonates. In the 6 h duration following DVET, 1
(2 %) neonate died from worsening sepsis; 12 (29 %) developed mild hypothermia (36.0 – 36.4 °C) during line placement
for DVET; and 2 (5 %) had transient bradycardia that spontaneously recovered. Within 48 h from DVET, 2 (5 %) had serum
potassium >6.5 mEq/L without any electrocardiography evidence of hyperkalemia and another 2(5 %) had serum sodium
>145 mEq/L and both recovered spontaneously. In comparison
to the ST group, significantly lesser neonates in the DVET
group had progression of thrombocytopenia [ST vs. DVET:
30 (71) vs. 14 (37); RR (95 % C.I): 2.1 (1.3–3.3); p 0.005]
and worsening of metabolic acidosis [ST vs. DVET: 12 (28)
vs. 3 (7); RR (95 % C.I): 3.9 (1.2–13); p 0.02] following DVET.
Discussion
Fig. 2 Kaplan – Meier Survival curve demonstrating cumulative survival
and time to death between both the groups
DVET has long been proposed and has been practiced in many
units as a therapeutic measure for severely septic neonates.
Many units perform DVET as a last resort to rescue sick and
severely septic neonates. However, its efficacy and safety as an
adjunctive therapy has been inconclusive despite its theoretical
benefits. The current study has shown a non-significant 21 %
reduction in the risk of mortality by 14 d from enrollment in
DVET group as compared to ST group. Similarly, early mortality (by 7 d from enrollment) and mortality by discharge also
showed a trend towards decrease in the DVET group in comparison to the ST group. No difference was observed in the time
to mortalitybetween the study groups. Amongst the individual
organ dysfunctions, cardiovascular dysfunction and hematological dysfunction benefitted the most by DVET followed by
renal dysfunction. However, no such benefit was observed with
112
Indian J Pediatr (February 2016) 83(2):107–113
Table 4 Effect of DVET on
hematologic, metabolic and
immunologic parameters
Characteristics
Pre DVET
Mean±SD
Post DVET
Mean±SD
MD
(95 % C.I.)
− 4 (− 7 to −1.1)
Pa
1
Hematocrit (in %)
45±7
49±6
2
3
Potassium (mEq/L)
Blood pH
4.9±1.1
7.26 ±0.1
4.6±1.1
7.31±0.1
0.12 (−0.1 to 0.7)
0.01 (−0.07 to −0.01)
0.16
0.002
4
Base deficit
5
6
7
8
Platelets (per mm3)
ANC (per mm3)
CRP (mg/L)
IgG (g/L)
−10±5
90,940±98,403
5585±9046
−7.4±5
79,121±55,809
3531±3053
0.7 (1.7 to 4.6)
11,818 (−24,911 to 48,548)
2053 (−2745 to 6851)
<0.001
0.5
0.4
49±39
5.7±2.2
32±35
8.1±1.5
9
10
IgA (g/L)
IgM (g/L)
0.3±0.2
0.22±0.16
1.06±0.3
0.76±1.5
− 0.8 (− 0.95 to −0.65)
− 0.53 (− 1.04 to −0.004)
<0.001
0.05
11
C3 (g/L)
0.6±0.18
0.83±0.2
− 0.22 (−0.32 to −0.12)
<0.001
17.9 (3.3 to 32)
− 2.5 (−3 to −1.4)
0.002
0.01
<0.001
DVET Double volume exchange transfusion; ST Standard therapy; MD Mean difference; ANC Absolute neutrophil count; CRP C-reactive protein; Ig Immunoglobulin; C3 Complement factor 3
a
Paired T test; all negative symbols indicate an increase following DVET
respiratory dysfunction. A significant improvement in the biochemical, immunological and acid base status was observed
following DVET.
From the developing countries, Sadana et al. reported a
significant improvement in survival in the DVET group in
comparison to no DVET group (50 vs. 5 %) in neonates with
sclerema [12]. Mathur et al. in an earlier study observed a
35 % absolute reduction in mortality in the DVET group in
comparison to no DVET group (70 vs. 35 %) [11]. However,
these two studies were done with a primary objective of analyzing the granulocyte function and immunoglobulin and
complement levels; both had a very high baseline mortality
risk (95 and 70 % respectively); had small numbers and suffered a high risk of bias. The current study had a baseline
mortality rate of 42 %. A direct comparison of the current
study with the above two may not be appropriate mainly due
to the significant improvement in the understanding of supportive care of sick neonates over the last two decades and an
increased awareness in effective and timely initiation of broad
spectrum antibiotics. Similar to the present results, a later
study done by Gunes et al. did not observe any significant
change in mortality between these groups [13].
As DVET may act by repleting the immunological competence of a sick neonate, and to understand the biological pathway of DVET, the authors also measured immunoglobulin
IgG, IgM and IgA along with complement factor 3 (C3), both
at baseline as well as after DVET and could observe a significant improvement in all these parameters post-DVET. A similar increase in the immunoglobulin levels following DVET
was observed by previous studies [7, 12, 13].
Instead of a clinician’s perception of sickness being a deciding factor for inclusion in the trial, the current study used
more objective criteria that mandated atleast one organ dysfunction to be present to classify sepsis as severe enough to be
enrolled in the trial [15]. None of the previous studies
have till now used such objective criteria for inclusion
nor have reported the organ dysfunction status after the
intervention. The level of sickness, as measured by SNAP II,
was similar at baseline between both the groups. A large number
of the study subjects [n=37 (45 %)] had SNAP II scores<20,
indicating a milder illness severity. Antibiotics and supportive
care formed the mainstay of standard therapy in severe sepsis. In
this study, both the groups received similar care during the
study period and the unit policy for starting and hiking up
of antibiotics and the choice of antibiotics remained unchanged during the study period.
Considering the multitude of complications involved in
more sick neonates, moribund neonates and those with severe
bleeding as well as refractory shock requiring more than 2
drugs were not included in the current trial. Similarly, extreme
low birth weight neonates were not included in this trial due to
the perceived higher risk of procedure related complications in
this population.
Safety of an elaborate procedure like DVET has been much
debated in the past. The authors took 6 h for mortality and 48 h
for other adverse effects as cut-offs to associate an event to
DVET. The lone subject in the DVET group who died had a
severe sickness level even at baseline (SNAP score of 35) and
had a relentless progression of multi-organ dysfunction.
Contrary to the belief, it was observed that the blood
pH and base deficit improved from 7.26±0.1 to 7.31±
0.1 and −10±5 to −7±5 (p<0.001) respectively, following
DVET. This improvement was observed despite the use of
slightly older blood with a mean donor blood pH of 6.75.
Previous studies had postulated that the transfused blood
was acidic predominantly due to the citrate content of
anticoagulant used [19, 20]. Tollner et al. [6] and Vain
et al. [7] also observed a similar improvement in
Indian J Pediatr (February 2016) 83(2):107–113
metabolic acidosis, with the former postulating that improvement in metabolic acidosis and oxygen requirement
might be associated with an overall improvement in the microcirculation. Similarly, only 2 (5 %) neonates in the DVET
group had serum potassium>6.5 mEq/L despite a donor blood
potassium of 10±4 mEq/L. This could be explained by the
complex in-vivo metabolic changes that occur at the level of
cell membrane of RBC in an environment of adequate ATP.
Previous studies have also reported a similar donor blood
potassium levels (5–27 mEq/L) [19, 20].
The current study has certain important strengths. First, this
was a rigorously conducted largest randomized controlled trial
with mortality as a primary outcome with a low risk of bias.
Secondly, objective and reproducible inclusion criteria were
applied to increase the generalizability of the results. Thirdly,
the level of sickness at baseline was objectively scored and
compared between the study groups with a validated severity
score like SNAP II with an excellent psychometric property.
Fourthly, serial assessment of improvement in organ dysfunction was done with objective criteria. Even though inability to
blind the intervention from the treating team as well as the
investigators was a limitation, mortality as an outcome would
have been least affected by this limitation. Moreover, all other
biochemical and immunological outcome variables were analyzed and reported in a blinded fashion.
To conclude, DVET showed a trend towards reduction in
mortality of 21 % in comparison to ST in severely septic
neonates of >1000 g birth weight. A significant improvement
was observed in the cardiovascular and hematological organ
functions following DVET. DVET was associated with a significant improvement in IgA, IgG, IgM, complement 3 and
metabolic acidosis in comparison to the standard therapy.
Thus, DVET is a safe procedure in severely septic neonates.
Contributions ASA: Conceptualized the study, collected the data
and drafted the initial manuscript and approved the final manuscript;
VS: Conceptualized and designed the trial, designed the data collection tool, supervised conduct of the trial, analyzed the data, critically
reviewed the manuscript and approved the final manuscript; PK: Conceptualized the study idea, supervised the design and implementation
of the trial, critically reviewed the data analysis and the manuscript
and approved the final manuscript; SMG: Assisted in designing the
trial and data collection, was involved in the study conduct, critically
reviewed the manuscript and approved the final manuscript; AJ:
Planned blood products administration, designed the data collection
tool, planned the preparation and administration of blood products,
critically revised the manuscript and approved the final manuscript;
AR: Contributed to the study design, conducted the analysis of biochemical and immunological parameters, reviewed the manuscript and
approved the final manuscript. PK will act as guarantor for this paper.
Conflict of Interest None.
Source of Funding None.
113
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Zaidi AKM, Huskins WT, Thaver D, Bhutta AZ, Abbas Z,
Goldmann AD. Hospital acquired neonatal infections in developing
countries. Lancet. 2005;365:1175–88.
Ganatra HA, Stoll BJ, Zaidi AKM. International perspective on
early onset neonatal sepsis. Clin Perinatol. 2010;37:501–23.
Tarnow-Mordi W, Isaacs D, Dutta S. Adjunctive immunologic interventions in neonatal sepsis. Clin Perinatol. 2010;37:481–99.
The INIS Collaborative Group. Treatment of neonatal sepsis with
intravenous immune globulin. N Engl J Med. 2011;365:1201–11.
Prod’hom LS, Choffat JM, Frenck N, Mazoumi M, Relier JP,
Torrado A. Care of the seriously ill neonate with hyaline membrane
disease and with sepsis (scleremaneonatorum). Pediatrics. 1974;53:
170–81.
Tollner U, Pohlandt F, Heinze F, Henrichs I. Treatment of septicemia in the newborn infant: choice of initial antimicrobial drugs and
the role of exchange transfusion. Acta Paediatr Scand. 1977;66:
605–10.
Vain NE, Mazlumian JR, Swarner OW, Cha CC. Role of exchange
transfusion in the treatment of severe septicemia. Pediatrics.
1980;66:693–7.
Lemos L. Exchange transfusion in treatment of sepsis. Pediatrics.
1981;68:471–2.
Narayanan I, Mitter A, Gujraj VV. A comparative study of the value
of exchange transfusion and blood transfusion in the management
of severe neonatal septicemia with sclrema. Indian J Pediatr.
1982;49:519–29.
Dalvi R, Rao S, Rangnekar J, Fernandez A. Exchange transfusions
in neonatal sepsis. Indian Pediatr. 1991;28:39–43.
Mathur NB, Subramaniam BKM, Sharma VK, Puri RK. Exchange
transfusion in neutropenicsepticemic neonates: effect on granulocyte functions. Acta Paediatr. 1993;82:939–43.
Sadana S, Mathur NB, Thakur A. Exchange transfusion in septic
neonates with sclerema: effect on immunoglobulin and complement
levels. Indian Pediatr. 1997;34:20–5.
Gunes T, Koklu E, Buyukkayhan D, Kurtoglu S, Karakucku M,
Patiroglu T. Exchange transfusion or intravenous immunoglobulin therapy as an adjunct to antibiotics for neonatal
sepsis in developing countries: a pilot study. Ann Trop Paediatr.
2006;26:39–42.
Mathur NB. Neonatal sepsis. Indian Pediatr. 1996;33:663–74.
Goldstein B, Giroir B, Randolph A. International pediatric sepsis
consensus conference: definitions for sepsis and organ dysfunction
in pediatrics. Pediatr Crit Care Med. 2005;6:2–8.
Variable block randomization software. Available online at: http://
www.randomization.com. Accessed on December 31 2011.
Amiel-Tison C. A method for neurologic evaluation within the first
year of life. In: Meyer L, editor. Current problems in pediatrics, VII,
no.1. Chicago: Year Book Medical Publishers; 1976.
Joliff CR, Cost KM, Stivrins PC, et al. Reference intervals for
serum IgG, IgA, IgM, C3, and C4 as determined by rate nephelometry. Clin Chem. 1982;28:126–8.
Mukherjee S, Marwaha N, Prasad R, Sharma RR, Thakral B. Serial
assessment of biochemical parameters of red cell preparations to
evaluate safety for neonatal transfusions. Indian J Med Res.
2010;132:715–20.
Lockwood WB, Leonard J, Liles SL. Storage, monitoring,
pretransfusion processing and distribution of blood components.
In: Roback JD, Grossman BJ, Harris T, Hilhyer CD, editors.
Technical manual. Bethedsa: American Association of Blood
Banks; 2011. p. 277.